I assume that an accurate estimation of training load is a crucial variable for a succesful static or AI-based training plan.
My swim load seems unreasonably high, compared to my runs, rides, and cross-country ski sessions and makes me hesitant of your product.
For example, a 34 minute swim (50% drills w/ fins and paddles) avg HR of 111 gets a load of 59, whilst a 38 min ride including sprints, a 6 min all-out CP-test, and avg HR 138 gets a load of 30.
What is the, physicological, explanation for this? Is the swim session load based purely on pace? If so, how do you take into account the “false high” speeds from drills? How does your model take into account or adapt to occasional/weekly sessions with inaccurate training load?
Hi @Stenfors,
Correct on the importance of load measurement. Typically errors arise from incorrect zone settings. I checked your swim settings and the workout you did and these don’t seem likely to be too far off. Remember that the load calculation comes from pace first, not heart rate. So fins and paddles could be making you go fast relative to your typical swim only pace. Unfortunately we can’t take into account false high speeds from drills at this point. However the difference of say 40 vs 50 load points is likely to be irrelavant in the big picture of adaptations needed session by session so to do so is likely splitting hairs. However if you were to miss the session 0 v 50 then that’s a different story.
Your cycling CP-test is impressive. However as shown in your profile, we haven’t yet been able to build a reliable model and require more data and longer rides. Your point is made and in fact we are working on logic to more appropriately weight larger high intensity bouts such as you did. Today we are using more of a duration biased load model but working towards one that leverages the critical power model as we allude to here. While I appreciate that you the user want this immediately, the reality is that these models take time, resource and testing before they can be appropriately be released. We are in the middle of that work now as we speak.
1 Like